mauricio carabar n , adri an de la garza and oth on m. moreno … · 2015-04-07 · global...
TRANSCRIPT
GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO
Mauricio Carabarın ∗, Adrian de la Garza† and Othon M. Moreno‡
Banco de MexicoThis draft: May 7 2015
Abstract. In the aftermath of the 2008 crisis, corporations in emerging market economiesincreased their issuance of debt securities in international financial markets under signifi-cantly better conditions. The undergoing process of normalization of monetary conditions,particularly in the U.S., poses potentially significant risks for these firms, especially as U.S.interest rates increase and exchange rates in emerging economies depreciate. This paper hastwo main objectives: (1) to describe the recent evolution of corporate financing in Mexicoand quantify the potential adverse effects on foreign corporate debt servicing to interest rateand exchange rate shocks; and (2) to assess how changes in global liquidity may affect boththe ability of large Mexican corporates to tap international financial markets and the accessof large and small and medium enterprises (SMEs) to domestic credit. The analysis uses anovel data set that combines credit-level data of the universe of all loans and lines of creditgranted by banks and non-banks to non-financial private firms in the country, the universeof all bond placements by these same companies, and balance sheet information of bond is-suing firms between 2003 and 2014. Our results suggest that the increase in global liquidityin the years following the Great Crisis significantly reduced international borrowing costsfor large debt-issuing firms in Mexico, which in turn explains the rise in foreign debt place-ments in recent years. We illustrate that, as large firms have turned away from domesticfinancing sources, commercial banks in Mexico have been able to funnel more resources toSMEs–a situation that may potentially reverse once international credit conditions becomemore restrictive and large firms return to domestic lending markets.
JEL classification: F32, F34, E44, G32
The contents of this article are the authors’ responsibility and do not necessarily reflect the views ofBanco de Mexico. The authors would like to thank seminar participants at the Second Federal Reserve Bankof New York-Banco de Mexico Workshop and the 6th BIS CCA Research Conference for helpful commentsand suggestions. We are grateful to Hector Reyes and Ana Laura Garcıa for their outstanding researchassistance. All remaining errors are ours.
∗E-mail: [email protected]. Av. 5 de Mayo #18, Centro 06059, Mexico, D.F. Phone: +52555237 3631.
†E-mail: [email protected]. Av. 5 de Mayo #18, Centro 06059, Mexico, D.F. Phone: +52555237 2412.
‡E-mail: [email protected]. Av. 5 de Mayo #18, Centro 06059, Mexico, D.F. Phone: +52555237 2476.
1
2 GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO
1. Introduction
The years after the global financial crisis of 2008-09 have been characterized by extremely low
policy interest rates in advanced economies and abundant liquidity in international markets. Under
these conditions, risk premia declined and international corporations, particularly from emerging
economies (EMEs), expanded their placements of securities overseas, thereby satisfying asset man-
agers’ and global investors’ appetite for higher returns. This “second phase of global liquidity,”as
dubbed by Shin (2013), has been characterized by investors’ search for yield (Turner 2014) and has
been a cause of major concern for both policymakers and academics because of its potential impli-
cations for cross-border financial stability. In particular, overall portfolio flows in EMEs may have
become more vulnerable to global financial conditions, especially since it has been shown that capi-
tal flows from institutional investors into EMEs tend to be relatively stable during normal times but
pull back more strongly and persistently under stress (IMF 2014). In addition, since international
debt placements by EMEs are typically in hard currencies, the rapid increase in EMEs corporate
bond issuance may have led to currency mismatches, which could increase their vulnerability to
external macroeconomic shocks (Shin 2013).
In broad terms, the Mexican case is no different from that of other EMEs. The ample liquid-
ity in international financial markets after the 2008-09 crisis was associated with more favourable
financing conditions abroad for Mexican corporations. Not only were Mexican firms able to secure
foreign-currency financing at lower interest rates than before, but they also managed to obtain
larger amounts at longer maturities. The average maturity of bonds issued by Mexican firms
overseas increased from around 10 years in 2004 to about 16 years in 2014. The dollar amount
per issue went from an average of about 0.5 billion to more than 1.0 billion in the same period.
Perhaps more importantly, this improvement in financing conditions did not occur in domestic
debt markets–at least not to the same extent–, which favoured the search for foreign financing my
Mexican corporations. Between 2010 and 2014, the gross medium-term bond issuance abroad by
Mexican non-financial corporations averaged 13 USD billion per year, which is roughly four times
the annual average observed between 2006 and 2009.
Moreover, this surge in bond issuance overseas has implied a shift in the financing structure of
Mexican corporations. In particular, given that only the country’s largest firms have access to
international debt markets, the rise in debt placements overseas seems to have freed up financial
resources from domestic markets, which banks have been able to funnel towards firms without
access to bond markets abroad, especially small and medium enterprises (SMEs). Indeed, while
commercial bank credit to large issuing companies contracted after the 2008-09 crisis, domestic
bank credit to SMEs has grown significantly and has been the most dynamic component of banks’
commercial loan portfolios.
The main objective of this paper is to better understand how changes in global financial con-
ditions have impacted firm financing in Mexico, particularly in terms of the volume and prices of
GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO 3
debt issues and bank loans. To this end, we break up the larger task at hand into three smaller
ones. First, we describe how financing to non-financial corporations in Mexico has evolved in the
last decade, focusing on the expansion of private bond issuance overseas and the terms at which
these bonds have been offered. Second, we quantify how changing global conditions during this
“second phase of global liquidity”may have affected the financing of Mexican firms in international
debt markets, particularly with respect to costs—proxied by the primary spread between an inter-
national bond and a domestic one of similar characteristics—and market activity measured by the
probability of observing a debt placement at any point in time. Third, we show that as Mexican
corporates increased their activity in international debt markets, they reduced their liabilities with
domestic banks, and estimate the extent to which this reduced bank lending to large companies may
have resulted in more credit to SMEs. In light of the current volatility in financial markets and the
expected changes in the monetary stance of advanced economies and global financial conditions in
the near future, these calculations allow us to assess both how vulnerable large Mexican corporates
may be to such changes and by how much bank lending to SMEs may decrease once international
financial markets dry up and large bond-issuing companies return to the domestic credit market.
In order to investigate these questions, we use information from original regulatory reports, in-
dividual bond prospectuses, and private data vendors to carefully assemble a proprietary data set
that combines credit-level data—including the universe of loans and lines of credit granted between
2003Q4 and 2014Q1 to non-financial private firms by banks and regulated non-banks domestically,
as well as all private bonds issued either domestically or internationally during this period—with
balance sheet information for issuing firms in Mexico. We classify firms as either large issuing firms,
large non-issuing firms or SMEs using information contained in regulatory reports on sales, number
of employees, and credit take-up. It is important to emphasize that issuing firms are examined
at the parent-company level, which implies that, for each parent company, all individual company
data were consolidated across subsidiaries.
Our results indicate that changes in global conditions in recent years have decreased the cost
of issuing debt externally relative to that paid by the same firm in the domestic market. This in
turn has increased the probability of a Mexican corporate issuing abroad. In particular, we estimate
that a 100 bp reduction in the external-internal debt issuing spread increases the probability of
issuing abroad by about 11%. In addition, our results suggest that, as Mexican corporates sought
foreign financing and decreased their credit take-up in the domestic loan market, SMEs benefited
from an increase in bank credit: a decrease of one MXN in new loans to large issuing firms increases
new loans to SMEs by 0.20 MXN. Combining these results, we estimate that a 100bp external-
internal debt issuing spread increase would reduce new bank loans to SMEs by about 4 billion
MXN, about 11% for the last observation in our sample. Although we perceive this to be a sizable
impact on bank credit to SMEs—with potential consequences to real sector activity—, it is pos-
sible that this crowding-out effect be mitigated, particularly if large issuing firms seek alternative
financing sources other than domestic bank loans, such as global bank loans, domestic debt issues
4 GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO
or trade credit. In addition, even if large issuing firms take up large amounts from the domestic
credit market, commercial banks may continue granting credit to SMEs given the significant push
that the recently-approved Financial Reform gives to joint loan provision between commercial and
development banks, and given low delinquency ratios and the relatively good credit history that
SMEs have built these last few years.
The rest of the paper is organized as follows. In Section 2, we present a brief review of the
literature. Section 3 describes the data collection process, provides some descriptive statistics and
describes some recent trends in corporate financing in Mexico, presenting some stylized facts with
regards to the recent surge of bond issuance in international markets by Mexican companies. Sec-
tion 4 discusses our estimation strategy and describes our main results regarding the effect of global
conditions on Mexican firms’ financing in international bond markets, particularly with respect to
costs and market activity. In Section 5, we present our findings about the substitution between
credit to debt-issuing firms and loans to SMEs. Finally, Section 6 concludes.
2. Review of the Literature
The concept of global liquidity has become increasingly popular in economic debate while re-
maining notoriously difficult to define. Most recent literature relates global liquidity to the ease
of financing conditions observed in global financial markets (BIS 2011). Empirical work produced
before the global financial crisis usually defined global liquidity in terms of the growth rate of mone-
tary aggregates from advanced economies, using either narrow or broad money measures (Baks and
Kramer 1999; D’Agostino and Surico 2009). Nevertheless, as financial integration has increased in
recent years–as evidenced by the presence of a highly integrated global financial cycle (Rey 2013)
and a larger amount of funds flowing from G4 economies to the rest of the world (Cerutti et al
2014)–, the issue of finding measures of global liquidity that are more adequate for investigating its
implications on cross-border real and financial linkages has garnered much attention.
In the aftermath of the 2008 crisis, the literature seemed to have shifted away from money-based
global liquidity measures towards the use of different types of indicators that attempt to proxy
liquidity conditions in international markets depending on the object of study. One such approach
has relied on quantity or volume indicators akin to monetary aggregates, typically employed in
studies that investigate potential financial sector vulnerabilities due to excess capital flows (Borio
et al. 2013; Bruno and Shin 2014; Cesa-Bianchi et al 2015). Another approach has been to combine
quantity and price variables, which helps in better gauging the ease of financing in global markets
(Domanski et al 2011; Chen et al 2012). Indeed, some have argued that global liquidity conditions
cannot be summarized by a single indicator (Eickmeier et al 2013) and that it is important to
differentiate between official and private liquidity–the latter having the stronger complementarity
with liquidity in domestic markets (Landau 2013). A large proportion of the recent empirical lit-
erature on global liquidity thus utilizes a quantity measure, such as cross—border credit, together
with indicators that relate to the willingness of private investors to provide funding, such as yield
GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO 5
spreads or risk appetite proxies like the VIX index (Domanski et al 2011; Chen et al 2012; Cerutti
2014).
Irrespective of the way it is measured, there is much agreement in the literature that global liquidity
has important effects on real and financial variables on both advanced and emerging economies,
particularly serving as “a vehicle for the numerous interactions and spillovers between domestic
monetary and financial policies”(Landau 2013). For example, Canova (2005) use a VAR model to
show that U.S. monetary shocks produce significant fluctuations in Latin American macroeconomic
variables, being the interest rate channel an important amplifier of these shocks on output, inflation
and nominal exchange rates in the region. Ruffer and Stracca (2006) employ a similar approach to
conclude that a global monetary policy shock has a positive effect on prices and output in the euro
area and Japan. More recently, Bayoumi and Bui (2011) show by means of event studies that the
U.S. monetary stimulus packages, particularly QE1, were associated with large reduction in emerg-
ing market yields and currency appreciations, and generally supported foreign activity. McCauley
et al (2015) use panel estimation techniques to illustrate how accommodative U.S. monetary policy
and cheap leverage have promoted credit growth globally. Finally, Azis and Shin (2013) studied the
effect of more ample global liquidity conditions in the years following the great financial crisis on
Asian economies. In particular, they find that rising non-core bank liabilities played an important
role in the transmission of global liquidity shocks to emerging Asia, and that the growing share of
foreign ownership of equity and securities in the region makes these markets more susceptible to
volatility and capital outflows.
This paper makes three key contributions to the literature on global liquidity and its impact on
domestic macrofinancial conditions. First, while most empirical and theoretical studies analyse the
potential spillover effects of excess capital flows via the banking sector, this analysis emphasizes
the role of international corporate debt markets. This is particularly relevant in the aftermath
of the 2008 crisis given the generalized boom in bond issuance in global markets by non-financial
firms in emerging economies. Second, our paper uses a comprehensive data set of the universe
of bond placements and individual commercial loans provided by the domestic banking sector to
study how global financial conditions changed the structure of credit and debt markets. To the
best of our knowledge, this is the first paper that uses microdata to this end. Third, in contrast to
most of the literature that relies on cross-country studies, panel regressions, and VAR models, our
data set allow us to employ a microeconometric approach to study the impact of global liquidity in
one particular case—that of the Mexican corporate sector. Although the Mexican case may differ
from that of other emerging economies, we believe that the results derived from this analysis may
provide some useful insights to academics and policymakers interested in understanding the po-
tential spillover effects that lax financial conditions globally may have had on developing economies.
Finally, our research also contributes to the literature regarding the liability structure of firms
6 GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO
and the ownership composition of debt portfolios. On the topic of firms’ liability structure, no-
table contributions include Colla et al (2012) and Morellec et al (2013), who present evidence of
higher diversification across multiple debt types for large rated firms, as well as for firms with
more growth options and higher bargaining power. These results are in line with Rauh and Sufi
(2010) who document debt heterogeneity in firms’ balance sheet debt, and important adjustments
in the underlying composition of corporate debt, even though no significant one year changes in
total debt are observed. Moreover, in what concerns what drives firms to issue bonds in foreign
markets, as opposed to domestic markets, Gozzi et al (2012) explore the use of international debt
market by firms using a database with information on new bond issues at a transactional level
including bonds from several countries. In particular, they find that international bond markets
complement the domestic markets, thus providing different financial services, other than issuing in
foreign currencies, not fulfilled in domestic markets. As shown in the next section, these findings
are corroborated by our description of the characteristics of domestic and foreign debt placements
by Mexican corporates.
3. Data description and recent evolution of corporate financing in Mexico
3.1. Data description.
In order to investigate the effect of global conditions on firm financing decisions in Mexico,
we gather a comprehensive dataset with credit-level and firm-level information on Mexican non-
financial private corporations for the period of 2003 Q4 to 2014 Q1. We complement these data
with a number of both global and domestic macroeconomic aggregates, including those aimed at
measuring the availability of financial resources in international markets. The remainder of this
section describes the sources and the main features of the data.
Credit-level data (individual bonds and loans)
We obtain data on the universe of corporate bonds issued either domestically or abroad by Mexi-
can non-financial private corporations between 2003 Q4 to 2014 Q1. Information for bonds issued
in international markets is retrieved from Bloomberg, which provides the main characteristics on
each issuance (e.g. issuer, issue and maturity dates, amount issued, yield at issue, credit rating,
currency, type of rate, periodicity of coupon payments). We fill in some missing data from firms
individual bonds prospectuses, and add a few other indicators from this source, such as the use
of proceeds from the debt placement. Tables A1 and A2 summarize the data. In total, there are
226 international bond issues by 53 non-financial Mexican firms during the sample period, most of
which are denominated in USD (87.6% of the number of bond issuances) and of maturities greater
than a year (99% of outstanding debt). The majority of these bonds were issued during the period
associated to the “second phase of global liquidity”—about two-thirds of bonds were issued after
the collapse of Lehmann in 2008 Q3. In terms of their characteristics, bonds issued internationally
after the crisis observed lower yields, larger amounts, and slightly longer maturities on average.
Although these are simple averages that do not necessarily reflect the greater heterogeneity of firms
GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO 7
participating in international bond markets, they are broadly consistent with what would be ex-
pected in an environment of ample liquidity and looser financial conditions.
For debt issued domestically, we collect individual bond-level data in a similar way. The main
sources of information are Indeval and Valmer, which provide financial data services on debt issued
by Mexican public and private entities. We fill in missing data using other sources, particularly
Bloomberg and individual bond prospectuses. To make these bonds comparable with those issued
abroad, we focus only on debt placements with maturities greater than a year. Tables A3 and A4
show some descriptive statistics. Our database consists of 316 bonds issued in domestic markets
during the sample period by 82 different firms. As expected, most of these bonds (87.3%) are
denominated in MXN, while 11% of these bonds are denominated in inflation-indexed units, UDIs.
Only four bonds in our sample were denominated in USD. In contrast to the marked increase in in-
ternational bond issuance after the 2008 crisis, domestic bond placements rose only slightly—about
53% of domestic private debt issues occurred after September 2008. In terms of their character-
istics, yields on domestic bonds decreased after the crisis, and the average amount per bond was
somewhat larger. However, the average maturity decreased slightly in the post-2008 period.
Finally, we also obtain information on the universe of loans provided by regulated financial inter-
mediaries in Mexico to non-financial private corporations using Banco de Mexico’s credit registry.
The data come from monthly regulatory reports (series R-04 C) administered by Banco de Mx-
ico and the National Banking and Securities Commission (CNBV for its Spanish acronym), which
provide transaction-level information on loans and lines of credit granted by regulated financial
intermediaries (banks and non-banks) to firms and individuals with entrepreneurial activity (or
personas fısicas con actividad empresarial in Spanish). We restrict our sample to non-financial
firms, which are identified by their RFC—a unique ID used for tax collection purposes. The sam-
ple period ranges from December 2003 to April 2014. For any given month, we observe individual
loans and lines of credit for each firm, as well as the main characteristics of the credit, including
interest rate, maturity, original amount, remaining balance, among others. As expected, bank
credit is much more widespread among Mexican firms than domestic or international bonds, at
least in terms of the sheer number of loans and lines of credit granted nationwide. In January 2014,
for example, there were 584,135 individual loans and lines of credit granted to 120,459 unique RFCs.
Using the above data, we classify firms as large companies or SMEs. Large companies are either
debt issuers (with participation in debt markets, either domestic or foreign) or non-debt-issuing
companies that meet at least one of the following conditions: (i) received a loan larger than 100
MXN million at any time during our sample period; reported having more than 250 employees; or
reported annual sales of more than 250 MXN million. In order to identify loans granted to large
issuing firms, it is important to acknowledge that many of these firms obtain financing through
their subsidiaries, which have their own RFCs. Thus, we collect information on the names of the
subsidiaries of all listed companies in Mexico, which they are required to disclose on their quarterly
8 GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO
financial reports. We used these names to find the RFC of each subsidiary and consolidate all data
across subsidiaries at the parent-company level.
Firm-level data
Balance sheet and general financial data on non-financial private corporations in Mexico come
mainly from financial reports of listed companies on the Mexican Stock Exchange (Bolsa Mexi-
cana de Valores, or BMV). These reports are published quarterly and their financial information
is electronically available in Banco de Mexico’s data system from 2007 onwards. We complement
this information to expand the sample period to begin in 2003 Q4 by manually collecting financial
data contained in individual reports retrieved directly from the BMV, as well as by using Econo-
matica, a private data provider. These efforts allow us to put together financial information on
139 different firms, whose dollar-volume of debt placements represents close to 90% of the total
dollar amount issued during the sample period. Individual firm characteristics include total assets,
which we employ as a measure of the size of the company. We use assets and sales denominated
in foreign currency to measure the “natural hedges”of these companies’ foreign liabilities. We also
collect information on firms’ profitability, liquidity, and short-term financing needs, such as ROA
and short-term assets and liabilities. We summarize these data in Table A5.
Figure 1 summarizes some of the information described above. During our sample period, 182 non-
financial private firms in Mexico obtained financing through either international or domestic debt
placements, or through the stock exchange. Out of these, 53 firms issued bonds in international
markets and 82 were active in domestic debt markets. We have balance sheet information on all
firms listed on the stock exchange. For non-listed firms the only data available is that related to
their activity in domestic credit markets, and their individual bond placements if they issued at
all.
3.2. Recent trends in corporate financing in Mexico.
The financing structure of firms in Mexico has undergone substantial changes after the global
financial crisis, highlighted by a significant increase in financing through external bond issuance
(Figure 2). In September 2008, external bond issuance represented 14% of total firm financing to
non-financial corporations in Mexico; by 2014 this figure has roughly doubled, which implies that
financing through external bond issuance has increased threefold in real terms during the period.
The growth rates of domestic financial sources have been positive but relatively much lower, while
external credit to these firms has decreased, congruent with the process of deleveraging by banks
in advanced economies following the crisis.
During the period that this shift in the liability structure of Mexican firms occurred, domestic
commercial bank credit to large firms with debt issuance – either abroad or domestically – began
to display a downward trend (Figure 3). It is important to note that this is only observed for issuing
firms: credit to large non-issuing firms has continued to grow in recent years, albeit at a slower pace
GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO 9
than credit to SMEs. From September 2008 to July 2009, the period when international financial
markets were most disrupted by the crisis, the trends described above were strikingly different:
during these months the amount of outstanding loans to large non-issuing firms stalled, loans to
SMEs decreased and loans to issuing firms surged.
Private bond placements in international markets by Mexican corporations increased significantly
in recent years. Annual gross bond issuance by private non-financial corporations in international
markets averaged 13 USD billion in 2010-2014, which contrasts with the average of 4 USD billion
recorded in 2002-2009. Net issuance averaged 10.1 USD billion in 2010-2014, up from 0.3 USD
billion in 2002-2009 (Figure 4). This surge in bond issuance has been mirrored by an increase in
the number of firms issuing in international markets. For example, in 2013 there were 21 private
non-financial corporations that issued bonds abroad, seven of them for the first time. In 2003, this
figure was much smaller, with only five firms issuing bonds abroad.
Greater access to international financial markets has implied greater diversification along several
characteristics for the set of Mexican firms issuing abroad. Firms from a wider variety of sectors
have been able to tap into these markets (Figure 5a). In 2003, around 70% of gross bond issuance
in international markets was from firms in telecommunication services. In contrast, this figure was
close to 30% in 2014, as sectors that previously had a very limited participation in these markets
have increased their issuance, such as materials and consumer discretionary and services sectors.
In addition, bonds with a wider variety of grades have been issued recently, congruent with the
entrance of new firms in the market and a greater appetite for risk and search for yield by interna-
tional investors (Figure 5b). Issuing terms for Mexican firms in international bond markets became
more favourable in the years following the 2008-09 crisis. Interest rates have become lower (Figure
6), while the average size and maturity of the issuances have increased. As Figure 6 shows, this
trend is true even when limiting the set of issuances to those made by firms with activity in both
domestic and international markets.
4. The effect of global liquidity on external bond issuance
The aim of this section is to quantify the potential impact of the “second phase of global liq-
uidity”on corporate bond issuance by Mexican firms in international markets. In particular, we
are interested in identifying how changes in financing conditions abroad associated with the ample
liquidity in financial markets may have affected the probability that a firm in Mexico issue an
international bond:
(1) Prob(Issuei,∗t = 1) = f(Spreadi,t , θ) + εi,t
The dependent variable is the probability that firm i issues a bond in international markets in
period t. This probability is a function of, among other things, the cost of placing this bond abroad
10 GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO
relative to what it would cost the firm to place a bond with similar characteristics (particularly
amount and maturity) in the domestic market. This cost is referred to as a spread or yield differ-
ential between the yield of an international bond and that of a domestic issue, as faced by firm i
at any given point in time. Defined this way, an increase in this spread would imply rising relative
costs of issuing debt abroad, which should in principle have a negative effect on a firm’s decision
to issue an international bond.
There are both theoretical and empirical challenges in estimating a model of the form we de-
scribed. First, there is no clear consensus in the literature as to what exactly drives corporate debt
financing in international markets (Gozzi et al 2012). The presence of frictions in financial markets
(at home or abroad) may affect firms’ financing decisions, particularly since debt securities and
loans may be characterized by different price and non-price attributes (Black and Munro 2010).
Non-price attributes could be related to several factors such as: (i) depth of market, as firms are
more likely to match their financing needs with lenders’ preferences in deeper markets, thus making
it easier to cumulatively raise more funds with a single or fewer issuances (Modigliani and Sutch
1966; Vayanos and Vila 2007); (ii) currency hedging, as issuances abroad may provide a useful nat-
ural hedge to their assets and cash flows denominated in foreign currency (Munro and Wooldridge,
2009); and (iii) funding diversification. Price attributes could be relevant because firms may also
issue abroad simply because the coupon rate of an international debt placement may become less
costly relative to that of a domestic bond.1
Besides these more theoretical issues, analyzing the particular case of Mexican firms presents
additional empirical challenges because bond market participation (either domestically or inter-
nationally) is relatively low, as shown in Section 3. The rather small number of individual bond
issues, as well as the fact that companies do not typically issue bonds in both markets around the
same time, hampers our ability to compare international and domestic debt financing costs for a
given firm. For example, during the period from 2003 Q4 to 2014 Q1, there were only 25 times
when a single firm issued bonds in both domestic and external markets in the same quarter. These
25 observations correspond to only eight different corporations. The interest rate differential for
a given period across markets may only be calculated in a straightforward manner for these few
observations. This problem is particularly relevant since the global liquidity effect may operate
mainly through the price channel (Shin 2013). Also, the price component is observed only when a
firm issues a bond, which immediately raises a well-known selection bias problem that should be
addressed.
To mitigate these concerns, we construct a counterfactual individual issuer-level spread time series
as a function of global financial conditions, domestic macroeconomic conditions, as well as indi-
vidual firm and bond characteristics. To do this, we first estimate the primary market coupon
1More generally, this question is related to a literature that links capital flows in EMEs to the effect ofboth push and pull factors (Calvo, Leiderman and Reinhart 1993, 1996; Schularick and Taylor 2012).
GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO 11
rate that each firm would pay both internationally and abroad for a bond of a given amount and
maturity. We then take the difference of these yield rates to construct the spread.
To be more precise, we first estimate two separate equations of the following form:
(2) y{D,∗}i,t = f(Global/DomesticV arst, BondChars
{D,∗}i,t , F irmCharsi,t) + ε
{D,∗}i,t
where the dependent variable is the observed primary market coupon rate for a bond issuance of
firm i in period t, either in domestic or international markets (D, *). We restrict our data to
bond issues in MXN and UDIs for the domestic market (99% of domestic issues) and in USD for
international markets (88% of international issues) in order to facilitate the comparison of interest
rates across currencies—as explained in more detail below. In terms of the particular controls use
in this estimation, we include global financial conditions, domestic macroeconomic indicators, and
individual firm and bond characteristics.
In terms of the proxies that reflect global financial conditions, we aim to capture the effect of
global liquidity on primary market coupon rates through price and quantity indicators as discussed
in Section 2. In particular, we consider as volume measure the total credit to non-residents in USD
and euros 2, obtained from the BIS Global Liquidity selected indicators. We also include the VIX
index as a price-based indicator of global liquidity reflecting investor’s risk appetite, and thus prox-
ying investors’ willingness to provide funding (Eickmeier et al 2013). In addition, we incorporate
the U.S. corporate BBB/BAA-Treasury 10-year spread in order to capture financing conditions
particular to debt markets, especially since flucutations in this particular spread should closely
mimic financing conditions faced by emerging economies’ corporates in international markets as
they approximately belong in the same asset class.
Regarding domestic conditions, we control for domestic activity using the year-on-year real GDP
growth rate in Mexico, and include the 28-day interbank interest rate (TIIE 28) as a proxy for do-
mestic financing conditions. Individual firm characteristics include a binary variable that indicates
whether the firm has ever issued an investment grade bond in external financial markets, as well
as lagged values of the firm’s total assets and return over assets (ROA). The latter two attempt
to measure firm size and profitability, which should have an impact on financing costs. Finally, we
control for maturity-at-issue and original amount of the bond.
The results presented in Table 1 correspond to the model where the international bond inter-
est rate is the dependent variable. Estimation (1) includes only bond characteristics. Estimation
(2) incorporates our global liquidity measures. Estimation (3) adds domestic conditions, and finally
estimation (4) includes firm characteristics. We find that the amount issued has a negative effect
2McCauley et al (2015), for example, use the same measure, but restricted to USD.
12 GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO
on the interest rate of international bond issuances, while maturity is not statistically significant in
any of our estimations. We find that cross-border credit is associated with a negative sign and is
always significant, which suggests that the abundance of liquidity in financial markets has reduced
the cost of issuing abroad to Mexican firms. In contrast, the U.S. corporate BBB/BAA-Treasury
10 year spread is associated positively to the yield, which suggests that, as risk appetite in debt
markets decreases, financing costs for Mexican firms increase. The VIX index is not statistically
significant in any of the estimations, although this is likely due to the high correlation with the
U.S. corporate spread (94% correlation). Finally, we also find that our variables concerning firm
characteristics all have the expected signs: firm profitability in the previous period seems to have a
negative effect on the interest rate paid in external bond issuances, while being a firm that has is-
sued bonds with investment grade in international markets also has a negative and significant effect.
The results from the model where the domestic bond interest rate is the dependent variable are
reported in Table 2. These results are ordered in a similar fashion to the previous table, as we add
different sets of variables to the model. Global conditions are found to have qualitatively similar
effects than in the model for external bond issuances, although neither the U.S. corporate spread
nor the VIX are significant in estimations (3) and (4). The TIIE 28 is positive and statistically
significant, which, unsurprisingly, implies that as domestic financing conditions become tighter, the
interest rates paid by firms in domestic debt issuances also rises.
With these results in hand, we obtain fitted values from the models that incorporate bond and
firm characteristics, as well as global and domestic conditions. We use these fitted values to com-
pute the interest rate that a given firm would have paid at each point in time and in each market
(D, ∗) for a bond with standard characteristics (500 million USD and 1 year maturity):
(3) {y∗, yDMXN}i,t
Then, in order to compare prices across markets, we transform the rates of bonds issued domesti-
cally by adjusting for expected depreciation:
(4) yDUSD =S
F(1 + yDMXN )− 1
Where yDMXN is the primary market coupon rate of domestic bonds, yDUSD is the primary market
coupon rate of bonds issued abroad, S is the spot exchange rate and F is a 1 year forward exchange
rate.
Finally, we calculate the spread at each point in time and for each firm as:
GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO 13
(5) Spread = y∗ − yDUSD
We plot the counterfactual spread we calculated in Figure 7, while Figure 8 and 9 present the
individual external and internal coupon rates for selected firms, respectively. The obtained spread
displayed a substantial increase at the height of the global financial crisis, when conditions in in-
ternational financial markets were most severely disrupted. Afterwards, there is a clear downward
trend in the spread, which implies that the relative cost of issuing abroad for Mexican firms has
been decreasing during the aftermath of the global financial crisis. This is congruent with the
increase of bond issuances in international markets by Mexican firms and the search for yield by
global investors during the period.
Having estimated this (counterfactual) spread for each individual firm, we then estimate how
cheaper financing conditions abroad may affect international bond issuance through a reduced-form
binary choice model as in equation (1) above. Our results are summarized in Table 3. Estimation
(1) controls only for whether a firm has issued in external markets before. Estimation (2) adds our
measure of global cross-border credit in hard currencies. Estimation (3) adds additional domestic
controls, such as domestic bank lending capacity.3 Estimation (4) adds other firm characteristics,
such as the ratio of short-term assets to short-term liabilities in order to control for firms’ financ-
ing needs. Using this last estimation, evaluating at means, we find that a 100bp reduction in the
foreign-domestic debt issuing spread increases the probability of issuing abroad by 10.5%. We also
find that having past experience in international debt markets is significant and has a positive im-
pact on the probability of issuing abroad across all estimations. Finally, our measure of cross-border
credit is not significant in any estimation, which suggests that most of its effect arises through prices.
5. The effect of global liquidity on domestic credit
In this section, we investigate how changes in global financial conditions have affected the do-
mestic credit market in Mexico. In particular, we are interested in examining whether the increase
in bond issuance in international debt markets by large debt issuing firms has reduced their par-
ticipation in domestic banks loan portfolio, thus freeing up resources that could then be allocated
for financing SMEs. Our motivation for estimating whether changes in global financial conditions
have made an impact on loans to SMEs stems from the stylized facts presented in Section 3 and
3 We use total obligations net of repo and derivatives operations since the banks’ balance sheets inMexico include transactional accounts used to record these types of operations that could overestimate theavailable loanable funds. In particular, we use the following definition for lending capacity::
LendCapn = TOg +max{0, derl − dera}+max{0, repol − repoa};where LendCapn is lending capacity, TOg is gross total obligations excluding derivatives and repo operations,derl and repol are the derivatives and repo operations on the liability side of the balance sheet, and similarlyfor the asset side.
14 GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO
our regression results in Section 4. First, we showed evidence of a change in the liability structure
of debt issuing firms—some of the largest in the country—consisting of a rapid increase in external
debt issuance coupled with a downward trend in outstanding volume of loans granted to these
firms. In Figure 10 we plot the share of external debt on total financing of non-financial corpora-
tions in Mexico that are either listed on the stock exchange or have participated in debt markets
throughout our sample period. This share has exhibited an increasing trend since September 2009,
growing from representing 39.6% of total financing to 64.3%. At the same time, loans to other
firms without participation in debt markets have displayed positive growth rates, particularly loans
to SMEs. Finally, our estimations in Section 4 showed that the increase in global liquidity has
implied a decrease in the relative price of issuing abroad that a given firm faces, which in turn has
increased the probability of issuing international markets.
Finding evidence of this crowding-in effect of loans to SMEs in Mexico is relevant for several
reasons. SMEs represent a large share of economic activity in Mexico: they are estimated to ac-
count for 72% of total employment in the formal sector and their economic activity is estimated
to represent 34% of GDP as of 2013 (Plan Nacional de Desarrollo, 2013). These figures are largely
consistent with the findings of Ayyagari, Beck and Demirguc-Kunt (2007), who report that SMEs
account for around 60% of manufacturing employment in a large sample of developed and devel-
oping countries. However, SMEs typically experience greater difficulty in obtaining resources from
the financial sector (Beck, Demirguc-Kunt, Laeven, and Maksimovic, 2006). The main theoretical
argument for explaining this phenomenon is that SMEs are often more informationally opaque than
larger firms, which in turn aggravates the agency problem inherent in corporate lending relation-
ships, and even more so for larger banks that tend to rely more on “hard”quantitative information
in their lending screening process (Berger and Udell 1998, Berger and Black 2011). Also, SMEs
are more likely to face financing constraints, since they usually have a smaller variety of financing
sources available for them.
Both of these mechanisms seem to be at work in Mexico, since SMEs tend to face higher interest
rates in the domestic commercial credit market than larger firms and also display slightly higher
delinquency rates. Additionally, SMEs in Mexico do not currently have access to debt markets.
According to the Credit Market Survey conducted by Banco de Mexico, around 23% of firms with
less than 100 employees received credit from commercial banks in December 2014, and none issued
debt. Most of these firms received financing through trade credit (around 65%). Finally, it is also
important to investigate the determinants of the volume of commercial bank credit to SMEs in
Mexico because of the relative scarcity of research devoted to the subject.4
In order to investigate whether there is indeed a crowding in effect of loans granted by commercial
banks to large debt-issuing companies on loans provided to SMEs, we estimate an equation of the
4 Lecuona (2009) and Fenton and Padilla (2012) provide general overviews of the topic.
GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO 15
following form:
(6) SMEt = f(Issuert, γ) + ut
where Issuer is a measure of loans granted by commercial banks in Mexico to debt issuing firms
and SME represents loans granted to SMEs. However, the main challenge in estimating the model
described above consists in overcoming a potential endogeneity problem. Since this is a reduced
form model, it is not difficult to argue that there may be other variables not included in the model
that are relevant in determining the amount of loans granted to both SMEs and debt issuing firms.
We attempt to address this issue by using an instrumental variable approach, informed by the
results in the previous section. In particular, we attempt to use variables that reflect the condi-
tions in international financial markets as instruments for the credit granted to issuing firms. In
particular, we use again the VIX index as a measure of risk appetite in financial markets, since it
reflects the availability of funding in international markets.5 This is potentially a valid instrument
since domestic credit to SMEs should not be directly affected by this variable since they themselves
do not have access to financing in international debt markets, and because we are also controling
for credit supply channels, like net bank obligations, and real demand channels, such as economic
growth in the country. At the same time, we have already showed how global variables affect the
participation in international markets of debt issuing firms, which in turn affects their demand for
loans in the domestic market. Furthermore, our estimations only use new loans granted in a given
period, which ameliorates concerns that the persistence inherent in using outstanding credit could
drive our results. Descriptive statistics on new loans for each type of firm is summarized in Table
A6. Finally, we control for the availability of resources in the domestic credit market by controlling
for our measure net bank obligations.
Thus, we estimate the following model:
SMEt =β1Issuert + β2LendCapt+Xγ + ut(7a)
Issuert = θ1V IXt + Zµ+ εt(7b)
where Issuer are new bank loans granted to debt issuing firms in a given period; SME are new
loans granted to SMEs in that same period; LendCap is our measure of domestic banks’ lending
capacity; X are additional variables meant to control for other macroeconomic aggregates that may
affect domestic credit market conditions, such as bank’s average funding costs and the economy’s
growth rate (measured by the IGAE index); and V IX represents the VIX index. The data for this
estimation has a monthly frequency and our sample ranges from December 2013 to April 2014.
5We run the same estimations using the US corporate spread as an additional instrument that capturesthe availability of funding in international financial markets. The results remain qualitatively unchanged.
16 GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO
Our results are summarized in Table 4. The first two regressions were estimated by OLS and
the coefficients associated with new loans to issuing firms are not significant. Once we implement
our instrumental variable strategy—in regressions (3) and (4)—, we find that this coefficient is
negative and statistically significant across all models. We interpret this result as evidence of a
crowding-in effect of loans to SMEs when debt issuing firms turn away from the domestic credit
market. Our measure of net bank obligations is positive and also statistically significant across
all models, which themselves differ as we use more macroeonomic variables as controls. The in-
strument we use is the monthly average of the VIX index, a standard measure of the market’s
expectation of stock market volatility. While perhaps the more natural variable to use as an instru-
ment would have been the firm-level interest rate spread we calculated before, we did not do it for
two reasons. First, we would have lost information for firms for which we were unable to build a
spread (e.g. firms with missing balance sheet data). Most importantly, the interest rate spread we
built had a quarterly frequency. As expected, the VIX has a significant and positive effect on new
loans granted to debt issuing firms in the first stage of our IV regressions, which suggests that, as
conditions tighten in international debt markets, large firms turn to the domestic credit market for
financing. Our preferred estimation is regression (4) in Table 4, which has additional controls for
macroeconomic aggregates. Using the coefficients associated with new loans to debt issuing firms,
we find that a decrease of 1 billion MXN in new loans to large issuing firms increases new loans to
SMEs by around 200 million MXN.
Finally, we are interested in using our results for estimating the potential impact of a tightening of
conditions in international financial markets on the volume of loans to SMEs. This is particularly
relevant in light of the expectation that there will be an increase in the policity interest rates of
some advanced economies–mainly the United States—in the coming months. If this increase in
the policy rates occurs and markets tighten, then our results suggest that Mexican debt-issuing
firms would face higher interest rates in debt markets abroad relative to the domestic market,
which would generate incentives for them to find financing in the domestic credit market. This
increase in the demand for loans by debt issuing firms could then crowd-out SMEs from the do-
mestic credit market, which could potentially hamper their productive activity as they have fewer
financial sources at their disposal.
In order to perform this calculation, we need to make some additional assumptions. We first
assume that gross bond issuance in external markets decreases in the same proportion as a de-
crease in the probability of issuing abroad by any firm, as estimated in Section 5. Furthermore, we
assume that these debt issuing firms that are unable or unwilling to place bonds in international
substitute completely their financing needs for domestic bank loans. That is, the decrease in debt
issuance abroad is symetricall to their increase in loans obtained through the domestic credit mar-
ket. This would mean that their position in other debt markets –such as the external credit market,
domestic debt market or trade credit—remains unchanged. Under these assumptions, combining
GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO 17
our results from our preferred estimations in Table 4, we find that an increase in 100bp in the
interest rate spread would translate into a crowding-out of new loans to SMEs of approximately
4 MXN billion, approximately 11% of the last observation in our sample. In turn, an increase of
1 standard deviation in the interest rate spread—equivalent to a surge of 159 basis points—would
lead to a crowding-out of new loans to SMEs of 6.4 MXN billion (17% of the last observation in
the sample).
18 GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO
6. Conclusions
In this paper, we studied how the abundance of global liquidity in the aftermath of the global
financial crisis affected firm financing in Mexico. We gathered both firm-level and credit-level infor-
mation in order to study how this change in global conditions had an effect across several markets
by changing Mexican firms’ financing structure. We found that the increase in global liquidity
seems to have reduced the cost of international borrowing for large debt-issuing firms in Mexico
relative to the cost of issuing bonds domestically, which in turn may explain the significant rise in
international debt placements in recent years. We also found evidence that shows that firms with
access to international bond markets—which are Mexico’s largest firms—changed their financing
structure during the period, substituting domestic financial sources for external sources. Among
domestic sources, we found that the amount of credit channeled through domestic financial inter-
mediaries displayed a downward trend during the period. Our estimations suggest that, as large
firms have turned away from the domestic credit market, banks’ capacity to funnel domestic credit
resources to SMEs has increased, generating a crowding-in effect on these firms. This is a particu-
larly relevant finding given that this situation may be reversed once international credit conditions
become more restrictive and large firms return to domestic lending markets.
It is important to bear in mind that this crowding-out effect may be mitigated if large issuing
firms seek alternative financing sources other than domestic bank loans (e.g., international bank
credit, domestic debt issues, trade credit). Additionally, credit conditions may soon become more
restrictive in domestic markets as well. Finally, it is possible that banks continue granting credit
to SMEs even if large firms return to the market, particularly if SMEs have displayed good credit
records since their entrance.
GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO 19
References
[1] Avdjiev, S., McCauley, R. and McGuire, P. (2012). “Rapid credit growth and international
credit:challenges for Asia”. Bank for International Settlements, 377.
[2] Ayyagari, M., Beck, T., and Demirguc-Kunt, A., 2007. “Small and Medium Enterprises across the
Globe”. Small Business Economics 29: 415-434.
[3] Azis, I.J. and Shin, H.S. (2013) “How do global liquidity phases manifest themselves in Asia?”. Asian
Development Bank.
[4] Baks, K., and Kramer, C. (1999). “Global liquidity and asset prices: Measurement, implications, and
spillovers”. International Monetary Fund 99-168.
[5] Bayoumi, T. A. and Bui, T. (2011). “Unforeseen Events Wait Lurking: Estimating Policy Spillovers
from US to Foreign Asset Prices”. International Monetary Fund, 11.
[6] Beck, T., Demirguc-Kunt, A., Laeven, L., and Maksimovic, V. (2006). “The determinants of financing
obstacles”. Journal of International Money and Finance, 25: 932-952.
[7] Berger, A. N., and Black, L. K. (2011). “Bank size, lending technologies, and small business finance”.
Journal of Banking and Finance, 35: 724-735.
[8] Berger, A. N., and Udell, G. F. (1998). “The economics of small business finance: The roles of private
equity and debt markets in the financial growth cycle”. Journal of Banking and Finance, 22: 613-673.
[9] Black, S. and Munro, A. (2010). “Why issue bonds offshore?”. Bank for International Settlements, 52.
[10] Borio, C., McCauley, R. and McGuire, P. (2011). “Global credit and domestic credit booms”. BIS
Quarterly Review, 3: 43-57.
[11] Borio, C. and Zhu, H. (2012). “Capital regulation, risk-taking and monetary policy: a missing link in
the transmission mechanism?”. Journal of Financial Stability, 8 236-251.
[12] Calvo, G. A., Leiderman, L. and Reinhart, C. M. (1993). “Capital inflows and real exchange rate
appreciation in Latin America: the role of external factors”. Staff Papers-International Monetary Fund,
40: 108-151.
[13] Calvo, G. A., Leiderman, L. and Reinhart, C. M. (1996). “Inflows of Capital to Developing Countries
in the 1990s”. The Journal of Economic Perspectives, 10: 123-139.
[14] Canova, F. (2005). “The transmission of US shocks to Latin America”. Journal of Applied econometrics,
20: 229-251.
20 GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO
[15] Cerutti, E., Claessens, S., and Ratnovski, L. (2014). “Global Liquidity and Drivers of Cross-Border
Bank Flows”. International Monetary Fund 14-69.
[16] Cesa-Bianchi, A., Cespedes, L. F., and Rebucci, A. (2015). “Global liquidity, house prices and the
macroeconomy: evidence from advanced and emerging economies”. Bank of England, 522.
[17] Chen, M. S., Liu, M. P., Maechler, A. M., Marsh, C., Saksonovs, M. S., and Shin, M. H. S. (2012).
“Exploring the dynamics of global liquidity”. Andrews McMeel Publishing, 12-246.
[18] Chung, K., Lee, J. E., Loukoianova, E., Park, H., and Shin, H. S. (2014). “Global liquidity through the
lens of monetary aggregates”. International Monetary Fund, 14.
[19] Colla, P., Ippolito, F. and Li, K. (2012). “Debt specialization”. The Journal of Finance, 68: 2117-2141.
[20] Bank for International Settlement, Committee on the Global Financial System. (2011). “Global
Liquidity-Concept, Measurement and Policy Implications”.
[21] D’Agostino, A., and Surico, P. (2009). “Does global liquidity help to forecast US inflation?”. Journal
of Money, Credit and Banking, 41:479-489.
[22] Domanski, D., Fender, I. and McGuire, P. (2011). “Assessing global liquidity”. BIS Quarterly Review,
4: 57-71.
[23] Eickmeier, S., Gambacorta, L. and Hofmann, B. (2014). “Understanding global liquidity”. European
Economic Review, 68: 1-18.
[24] Fenton, R., and Padilla, R. (2012).“Financiamiento de la banca comercial a micro, pequeas y medianas
empresas en Mexico”. Serie Estudios y Perspectivas, 135: 1-66.
[25] Gobierno de Mexico. Plan Nacional de Desarrollo 2013-2018.
[26] Gozzi, J. C., Levine, R., Peria, M. S. M. and Schmukler, S. L. (2012). “How firms use domestic and
international corporate bond markets”. National Bureau of Economic Research, 17763.
[27] International Monetary Fund. (2014). “Global Financial Stability Report- Moving from Liquidity- to
Growth Driven Markets”.
[28] Landau, J. P. (2013). “Global liquidity: Public and private”. Manuscrito para la Reunin Jackson Hole
2013.
[29] Lecuona, R. (2009). “El financiamiento a las pymes en Mexico, 2000-2007: el papel de la banca de
desarrollo”. Serie Financiamiento del Desarrollo, 207: 3-50.
GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO 21
[30] McCauley, R. N., McGuire, P., and Sushko, V. (2015). “Global dollar credit: links to US monetary
policy and leverage”. Bank for International Settlements, 483.
[31] Modigliani, F., and Sutch, R. (1966). “Innovations in interest rate policy”, American Economic Review,
56(1/2): 178-197.
[32] Morellec, E., Valta, P., and Zhdanov, A. (2013). “Financing Investment: The Choice between Bonds
and Bank Loans”.
[33] Munro, A., and Wooldridge, P. (2009). “Motivations for swap-covered foreign currency borrowing”.
Bank for International Settlements, 2-42. Preliminary draft.
[34] Rauh, J. D. and Sufi, A. (2010). “Capital structure and debt structure”. Review of Financial Studies,
23: 4242-4280.
[35] Rey, H. (2013). “Dilemma not trilemma: the global financial cycle and monetary policy indepen-
dence”. Proceedings of the Federal Reserve Bank of Kansas City Economic Symposium at Jackson Hole.
[36] Ruffer, R. and Stracca, L. (2006). “What is global excess liquidity, and does it matter?”European
Central Bank, 696.
[37] Shin, H. S. (2013). “The second phase of global liquidity and its impact on emerging economies”. In
Keynote address at the Federal Reserve Bank of San Francisco Asia Economic Policy Conference.
[38] Schularick, M., and Taylor, A. M. (2012). “Credit Booms Gone Bust: Monetary Policy, Leverage Cycles
and Financial Crises, 18702008”. American Economic Review, 102: 1029-1061.
[39] Turner, P. (2014). “The global long-term interest rate, financial risks and policy choices in EMEs”.
Bank for International Settlements, 441.
[40] Vayanos, D., and Vila, J. (2007). “A preferred habitat model of the term structure of interest rates”.
NBER Working paper.
22 GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO
Tables and Figures
Table 1. Dependent variable: International bond interest rate
(1) (2) (3) (4)Bond characteristicsMaturity -0.033 -0.041 -0.039 0.012
(0.052) (0.060) (0.058) (0.038)Amount (ext) -1.405*** -1.347*** -1.355*** -0.229
(0.227) (0.276) (0.271) (0.195)Global conditionsCross-border credit -0.346** -0.397* -0.499***
(0.134) (0.228) (0.122)US corporate spread 0.470 0.717 1.108*
(0.514) (0.590) (0.649)VIX 0.027 0.008 -0.016
(0.051) (0.057) (0.050)Domestic conditionsGDP Mex (yoy) 0.047 0.052
(0.082) (0.090)TIIE 28 -0.056 -0.059
(0.192) (0.128)Firm characteristicsInvestment grade (ext) -2.146***
(0.557)Total assets (lag) -0.236
(0.158)Profitability (lag) -0.073**
(0.033)
Observations 157 157 157 97Adj. R-squared 0.490 0.534 0.529 0.658
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1
GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO 23
Table 2. Dependent variable: Domestic bond interest rate
(1) (2) (3) (4)Bond characteristicsMaturity 0.124*** 0.142*** 0.125*** 0.112***
(0.020) (0.028) (0.028) (0.029)Amount (int) -0.564*** -0.472*** -0.470*** -0.250*
(0.161) (0.115) (0.105) (0.138)Global conditionsCross-border credit -0.790*** -0.227*** -0.241**
(0.088) (0.073) (0.092)US corporate spread 1.542*** 0.571 0.504
(0.336) (0.356) (0.483)VIX -0.123*** -0.031 0.000
(0.040) (0.033) (0.040)Domestic conditionsGDP Mex (yoy) 0.011 0.041
(0.050) (0.055)TIIE 28 0.612*** 0.564***
(0.091) (0.135)Firm characteristicsInvestment grade (ext) -0.645*
(0.347)Total assets (lag) 0.019
(0.131)Profitability (lag) -0.037
(0.023)
Observations 233 233 233 146Adjusted R-squared 0.138 0.463 0.548 0.525
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1
24 GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO
Table 3. Regression results: Probability of issuing abroad
(1) (2) (3) (4)Logit model
Issuance spread (fitted values) -0.107* -0.108* -0.118* -0.113*(0.058) (0.059) (0.064) (0.065)
Has issued in external market 0.788*** 0.698** 0.700** 0.688**(0.246) (0.278) (0.279) (0.279)
Cross-border credit 0.053 -0.055 -0.055(0.077) (0.271) (0.272)
Domestic conditions NO NO YES YESFirm characteristics NO NO NO YES
Observations 1,130 1,130 1,130 1,130Pseudo R-squared 0.0208 0.0215 0.0218 0.0233
Standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1
Table 4. Regression results: Effect of Global Liquidity on Domestic Credit
(1) (2) (3) (4)New loans to SME
OLS IVIssuer (new loan) -0.004 -0.009 -0.088*** -0.200***
(0.020) (0.027) (0.034) (0.065)Net lending capacity 46.868*** 42.260*** 47.413*** 51.422***
(2.161) (3.120) (2.200) (5.000)Bank’s avg. funding cost -0.631** 0.712
(0.315) (0.646)IGAE (Mex growth rate) -0.223* -0.641***
(0.130) (0.194)Large (new loan)
Constant Yes Yes Yes Yes
Observations 124 124 124 124F-Stat (first stage) 90.82 35.51Adj-R2 0.804 0.809 0.778 0.727
Standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1
We instrument Issuer (new loan) with the VIX index. VIX index is possitive and significantat p<0.01 in the first stage regressions of estimations (3) and (4). F-stat in these estimationsare 90.8 and 35.5, respectively.
GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO 25
Figure 1. Participation in (International/Domestic) Debt and Stock Markets byFirms in Mexico
Total 137
With external debt issuance 53
With internal debt issuance 82
Number of firms
18 firms
72 firms 24 firms
10
33
22
3
International Debt Market
Domestic Stock Market
Domestic Debt Market
Source: Banco de Mexico with information from BMV, Bloomberg and Indeval.
26 GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO
Figure 2. Financing to Non-Financial Corporations in Mexico.Index, 2008q3=100
0
50
100
150
200
250
300
350
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Domestic credit
Domestic bond issuance
External credit
External bond issuance
Source: Banco de Mexico.
GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO 27
Figure 3. Commercial Bank Credit to Non-Financial Corporations in Mexico byFirm Type1/.MXN billion
0
50
100
150
200
250
300
350
400
450
500
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Large firms, debt issuers
Large firms, non-issuers
Small and medium firms
Source: Banco de Mexico.1/ Only includes loans denominated in MXN.
28 GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO
Figure 4
(a) Gross Bond Issuance by MexicanCorporations in International Bond
Markets1/.USD billion
0
2
4
6
8
10
12
14
16
18
20
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
Non-investment grade
Investment grade
(b) Number of Mexican CorporationsIssuing Bonds in International
Markets1/2/
7 6
8
14
11
15
3
15 14
15
18
22
16
1 1 1
4
2
4
1
34
0 1
56
0
5
10
15
20
25
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
Source: Prepared by Banco de Mexico based on information from Bloomberg.1/ Includes bonds with maturity greater than one year issued in the external market by
non-financial private corporations.2/ The database has information since 1992, which means that new entrants for each year are
those that issued bonds in international markets in the corresponding period but had notdone so since 1992.The number of firms that match this criterion is represented by the white
figure inside each column.
GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO 29
Figure 5
(a) Gross Bond Issuance by MexicanCorporations in International
Markets by Sector.Percent
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
Energy Industrials
Consumer discretionary & Services Materials
Consumer staples Telecommunication services
(b) Gross Bond Issuance by MexicanCorporations in International
Markets by Grade1/.USD billion
0
2
4
6
8
10
12
14
16
18
20
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
BB
BBB
A
B
CCC-D
NR
Source: Prepared by Banco de Mexico based on information from Bloomberg.1/ Corresponds to the grade awarded to the issuer in the corresponding foreign market based
on Bloomberg Composite.
Figure 6
(a) Interest Rates of BondsIssued Abroad by Mexican
Corporations withInvestment Grade1/.
Percent
0
1
2
3
4
5
6
7
2003q4-2008q2
2010q3-2014q1
1 to 5 yrs
5 to 10 yrs
15+ yrs
10 to 15 yrs
(b) Average Maturity ofBonds Issued by Mexican
Corporations2/.Years, 3-year moving average
0
2
4
6
8
10
12
14
16
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
Domestic market
International market
(c) Average Maturity ofBonds Issued by Mexican
Corporations2/.Years, 3-year moving average
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
Domestic market
International market
Source: Prepared by Banco de Mexico based on information from Bloomberg and Indeval.1/ Includes only bonds denominated in USD.
2/ Includes only corporations that issued both in the domestic and foreign market.
30 GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO
Figure 7. Counterfactual Spread (International Minus Domestic Interest Rate).Percentage points
0
1
2
3
4
5
6
7
8
9
10
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Investment grade issuer
Non-investment grade issuer
Source: Prepared by Banco de Mexico based on information from Bloomberg.
GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO 31
Figure 8. Fitted and observed external coupons.Percent
0
4
8
12
2004
2004
2005
2006
2006
2007
2008
2008
2009
2010
2010
2011
2012
2012
2013
2014
AMX
0
4
8
12
16
2004
2004
2005
2006
2006
2007
2008
2008
2009
2010
2010
2011
2012
2012
2013
2014
CEMEX
0
4
8
12
16
2004
2004
2005
2006
2006
2007
2008
2008
2009
2010
2010
2011
2012
2012
2013
2014
GEO
0
4
8
12
16
2004
2004
2005
2006
2006
2007
2008
2008
2009
2010
2010
2011
2012
2012
2013
2014
VITRO
0
4
8
12
16
2004
2004
2005
2006
2006
2007
2008
2008
2009
2010
2010
2011
2012
2012
2013
2014
URBI
0
4
8
12
2004
2004
2005
2006
2006
2007
2008
2008
2009
2010
2010
2011
2012
2012
2013
2014
BIMBO
0
4
8
12
2004
2004
2005
2006
2006
2007
2008
2008
2009
2010
2010
2011
2012
2012
2013
2014
FEMSA
0
4
8
12
2004
2004
2005
2006
2006
2007
2008
2008
2009
2010
2010
2011
2012
2012
2013
2014
ALFA
0
4
8
12
2004
2004
2005
2006
2006
2007
2008
2008
2009
2010
2010
2011
2012
2012
2013
2014
TELEVISA
Source: Prepared by Banco de Mexico based on information from Bloomberg.
32 GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO
Figure 9. Fitted and observed internal coupons.Percent
0
4
8
2004
2004
2005
2006
2006
2007
2008
2008
2009
2010
2010
2011
2012
2012
2013
2014
AMX
0
4
8
2004
2004
2005
2006
2006
2007
2008
2008
2009
2010
2010
2011
2012
2012
2013
2014
CEMEX
0
4
8
2004
2004
2005
2006
2006
2007
2008
2008
2009
2010
2010
2011
2012
2012
2013
2014
GEO
0
4
8
2004
2004
2005
2006
2006
2007
2008
2008
2009
2010
2010
2011
2012
2012
2013
2014
VITRO
0
4
8
2004
2004
2005
2006
2006
2007
2008
2008
2009
2010
2010
2011
2012
2012
2013
2014
URBI
0
4
8
2004
2004
2005
2006
2006
2007
2008
2008
2009
2010
2010
2011
2012
2012
2013
2014
BIMBO
0
4
8
2004
2004
2005
2006
2006
2007
2008
2008
2009
2010
2010
2011
2012
2012
2013
2014
FEMSA
0
4
8
2004
2004
2005
2006
2006
2007
2008
2008
2009
2010
2010
2011
2012
2012
2013
2014
ALFA
0
4
8
2004
2004
2005
2006
2006
2007
2008
2008
2009
2010
2010
2011
2012
2012
2013
2014
TELEVISA
Source: Prepared by Banco de Mexico based on information from Bloomberg.
GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO 33
Figure 10. Share of external bond issuance in total financing for firms in ourmain sample.
Percent
0
10
20
30
40
50
60
70
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
34 GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO
Appendix
Table A1 Foreign bond issuances. Includes bonds issued in foreign markets by non-financial firms residing in Mexico. Issuances are considered investment grade when rated BBBor higher under BBG composite rating.
Full sample Pre 2008Q3 Post 2008Q3
Number of firms 53 32 40
All currenciesNumber of bond placements 226 77 149Investment grade 84 22 62Non-investment grade 142 55 87
USD onlyNumber of bond placements 198 73 125Investment grade 58 20 38Non-investment grade 140 53 87
Table A2 Foreign bond issuances: Main features. Includes bonds issued in foreignmarkets by non-financial firms residing in Mexico. Yield is the annualized primary marketcoupon rate. Amount is measured in millions of US dollars. Maturity represents maturity atissuance expressed in years.
Full sample Pre 2008Q3 Post 2008Q3Mean Std. Dev. Mean Std. Dev. Mean Std. Dev.
All currenciesYield 7.6 3.2 8.2 2.9 7.3 3.3Amount (millions) 376.6 348.9 308.6 257.4 411.8 383.9Maturity (years) 9.7 9.1 9.2 6.6 10.0 10.1
USD onlyYield 8.0 3.1 8.1 3.0 7.9 3.2Amount (millions) 340.4 323.2 302.0 258.9 362.8 354.5Maturity (years) 8.5 6.5 8.5 5.6 8.5 6.9
GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO 35
Table A3 Domestic bond issuances. Includes bonds issued in the domestic market bynon-financial firms residing in Mexico. Issuances are considered investment grade when ratedBBB or higher under BBG composite rating.
Full sample Pre 2008Q3 Post 2008Q3
Number of firms 82 43 61
All currenciesNumber of bond placements 316 146 170Investment grade 294 140 154Non-investment grade 22 6 16
MXN and UDI onlyNumber of bond placements 312 146 166Investment grade 291 140 151Non-investment grade 21 6 15
Table A4 Domestic bond issuances: Main features. Includes bonds issued in the domes-tic market by non-financial firms residing in Mexico. Yield is the annualized primary marketcoupon rate. Amount is measured in millions of US dollars.Maturity represents maturity at issuance expressed in years
Full sample Pre 2008Q3 Post 2008Q3
Mean Std. Dev. Mean Std. Dev. Mean Std. Dev.All currenciesYield 8.2 2.3 9.4 2.1 7.1 1.9Amount (millions) 129.0 129.7 115.0 108.3 141.0 144.8Maturity (years) 7.5 6.4 7.9 6.9 7.1 5.9
MXN and UDI onlyYield 8.2 2.3 9.4 2.1 7.2 1.9Amount (millions) 126.7 125.5 115.0 108.3 137.0 138.3Maturity (years) 7.5 6.4 7.9 6.9 7.1 6.0
36 GLOBAL LIQUIDITY AND CORPORATE FINANCING IN MEXICO
Table A5 Firm characteristics. All variables measured in millions of MXN pesos, exceptROA (Return over assets) which is the ratio of net income over total assets.
Mean Std. Dev. Min Max N
Total assets 42,750.8 116,775 5.48 1,383,459.0 3,955Short-term assets 13,546.5 30,197 0.29 400,008.3 3,955Short-term liabilities 9,622.3 27,654 7.27 376,741.9 3,955ROA 0.2 0 -2.34 1.1 3,902Assets in foreign currency 6,556.3 3,6610 0.00 573,925.4 2,128Outstanding foreign debt (bond issuance only) 2,295.4 15,989 0.00 425,877.9 6,552Outstanding domestic debt (bond issuance only) 1,274.0 4,143 0.00 64,344.0 6,552
Table A6 Domestic credit by type of firm. Debt issuer are non-financial firms residing inMexico that have issued bonds in either the domestic or international debt markets throughoutthe sample period. Large non-issuer are non-financial firms residing in Mexico that have beengranted a loan of over 100 million MXN pesos at any point in time during our sample period,have 250 or more employees, or have reported sales of over 250 million MXN pesos, whileremaining inactive in debt markets. SMEs are non financial firms residing in Mexico that donot fulfill any of the previous criteria. The volumen of new loans represents loans grantedby regulated financial intermediaries residing in Mexico to private non-financial firms withresidence in Mexico, expressed in billions of MXN pesos.
Volume of new loansMean Std. Dev. Min Max
TotalDebt issuer 39.3 15.2 18.5 9 .3Large, non-issuer 52.7 17.3 26.3 99.8SME 22.7 8.0 10.3 43.1
Pre 2008Debt issuer 37.2 15.3 18.5 75.8Large, non-issuer 38.5 8.1 26.3 63.0SME 16.5 4.4 10.3 30.7
Post 2008Debt issuer 41.0 15.0 19.0 97.3Large, non-issuer 64.2 14.0 38.2 99.8SME 27.8 6.7 15.8 43.1